1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
|
package Cache::Ref::CLOCK;
BEGIN {
$Cache::Ref::CLOCK::AUTHORITY = 'cpan:NUFFIN';
}
BEGIN {
$Cache::Ref::CLOCK::VERSION = '0.04';
}
# ABSTRACT: CLOCK cache replacement algorithm
use Moose;
use namespace::autoclean;
extends qw(Cache::Ref);
with qw(Cache::Ref::CLOCK::Base);
has k => (
isa => "Int",
is => "ro",
default => 1,
);
sub _hit {
my ( $self, $e ) = @_;
my $k = 0+$self->k; # moose stingifies default
$_->[0] = $k for @$e;
}
__PACKAGE__->meta->make_immutable;
__PACKAGE__;
# ex: set sw=4 et:
__END__
=pod
=encoding utf-8
=head1 NAME
Cache::Ref::CLOCK - CLOCK cache replacement algorithm
=head1 SYNOPSIS
my $c = Cache::Ref::CLOCK->new(
size => $n,
k => $k,
);
=head1 DESCRIPTION
This algorithm is provides a second chance FIFO cache expiry policy using a
circular buffer.
It is a very well accepted page replacement algorithm, but largely for reasons
which are irrelevant in this context (cache hits don't need to be serialized in
a multiprocessing context as they only require an idempotent operation (setting
a bit to 1)).
=head1 ATTRIBUTES
=over 4
=item size
The size of the live entries.
=item k
This is the initial value given to all hit entries.
As the hand moves through the circular buffer it decrements the counters.
The default is C<1>, providing semantics similar to a second chance FIFO cache.
Larger values of C<k> model LRU more accurately.
This is pretty silly though, as L<Cache::Ref::LRU> is probably way more
efficient for any C<k> bigger than 1.
=back
=head1 AUTHOR
Yuval Kogman
=head1 COPYRIGHT AND LICENSE
This software is copyright (c) 2010 by Yuval Kogman.
This is free software; you can redistribute it and/or modify it under
the same terms as the Perl 5 programming language system itself.
=cut
|
